Determination of Dialysis Access Patency Using 2D Angiographic Images

Abstract

Quantification of intra-access blood flow rate (BFR) is an important parameter for evaluating hemodialysis access patency in end-stage renal disease patients. In this study, we use bolus tracking method to compute BFR from clinical images obtained using digital subtraction angiography and evaluate the quantification accuracy versus thermodilution (TD) flow measurement using four different bolus transit time algorithms. We found that the cross-correlation method had the best overall accuracy with mean quantification error of 27+/-13% while all other methods had error >44%. The calculated BFR was dependent on frame rate and acquisition duration and the highest computational accuracy (83-115% of TD flow) was obtained for images with complete bolus flow-in and wash-out phases. The results of this investigation will be useful for developing flow measurement tool and preparing custom acquisition protocol for angiography suite.